Chips, Energy, Storage — The Three Pillars of AI Infrastructure: Who Will Rise First, Who Will Surge the Most, and Who Still Has Room to Catch Up?
Original author: Changan, Biteye

In November last year, Justin Sun posted a tweet:

If you view this statement as an industry assessment rather than a viral catchphrase, you’ll realize upon reflection that:

These three lines represent the most realistic profit path for AI market trends.

What would the result be if you had bought U.S. stocks related to data storage after that tweet was posted?

· Micron: +214%

· Seagate: +180%

· Western Digital: +190%

· SanDisk: +552%

Let’s break this article down along these three lines:

Why will AI first benefit chips, then create an energy bottleneck, and ultimately increase long-term storage demand? Which assets have already emerged in this structure?

I. Chip: What's being realized first from the AI boom is not narrative, but orders.

What was ignited first by AI is not the application layer, but the underlying computing power.

Whether training large models, or performing everyday inference, Agent calls, or multimodal processing, the first step is always to get the computations running—and these computations ultimately depend on GPUs, HBM, high-speed interconnects, and advanced process technologies.

In other words, the growing demand for AI will not first ripple through later stages, but will instead become a more immediate reality:

More chips, stronger chips, chips with higher bandwidth.

This is why AI demand is first reflected in the chip sector.

Industry data has clearly documented this: for fiscal year 2026, NVIDIA's revenue increased by 65% year-over-year, indicating that demand for high-performance computing chips continues to grow.

Core Computing Power Layer: NVIDIA (NVDA), AMD, Broadcom (AVGO), TSMC (TSM)

Domestic computing power layer: Hygon Information (688041.SH), Cambricon (688256.SH), etc. Hygon Information is one of the leading domestic enterprises in x86 server CPUs, with revenue of RMB 9.162 billion in 2024, a 52.4% year-over-year increase.

Semiconductor equipment layer: ASML, Applied Materials (AMAT), Lam Research (LRCX). The stock price of ASML, the leading lithography machine manufacturer, in the form of U.S. ADRs, has reached a new all-time high at the start of 2026, with a single-day gain of over 8% on January 2 and a cumulative rise of 27% since the beginning of the year; Lam Research has risen by as much as 30% since the start of the year; Applied Materials has gained as much as 28%. The stock prices of these three major semiconductor equipment giants have significantly outperformed the S&P 500 Index.

The chip sector was the earliest to rally and experienced the largest gains in this AI-driven market surge. NVIDIA, as the leader, has seen a cumulative gain of over 1,000% since early 2023. Equipment makers continued to hit new highs in early 2026 and remain in a strong upward cycle.

Citigroup released a research report predicting that the global semiconductor equipment sector will enter a "Phase 2 bull market cycle," with the primary focus for chip stocks in 2026 on ASML, Lam Research, and Applied Materials.

II. Energy: As AI grows larger, the bottleneck shifts from chips to electricity

No matter how many chips you have, without power, they won't run.

Buying chips is just the beginning; running large models, data centers, and inference services over the long term requires continuous power supply, along with additional cooling and heat dissipation demands.

Traditional data centers typically have a power capacity of 5 to 15 kilowatts per cabinet, while AI data centers have significantly increased to 50 to 100 kilowatts, creating a completely different scale of power consumption and cooling demands.

The IEA's analysis this year notes that data center electricity consumption is projected to increase to approximately 945 TWh by 2030, roughly doubling from current levels, with AI being the primary driver. The U.S. Department of Energy has also explicitly stated that the growing electricity demand from data centers is placing significant strain on regional power grids.

Gas turbines: GE Vernova (GEV): Gas turbine orders have surged, with full-year 2025 orders reaching $59 billion and backlog growing to $150 billion; management has raised its 2026 revenue guidance to $44–45 billion.

Independent power producer: Constellation Energy (CEG): The largest zero-carbon power operator in the U.S., with nuclear assets directly signing long-term power purchase agreements with major tech companies;

Vistra (VST): Combines nuclear and natural gas assets, with a mid-point 2026 EBITDA guidance approximately 30% higher than 2025.

Uranium resources: Cameco (CCJ): The world's largest publicly traded uranium miner and a primary beneficiary of the nuclear power revival.

GE Vernova's stock has risen 167% over the past year. The 52-week low was $408, and it reached a high of $1,181, nearly doubling in value within this range.

Constellation Energy reached an all-time high in 2025, then retraced approximately 28% from its peak due to regulatory policy fluctuations, and is currently at a relatively low level.

Vistra remains strong overall, with long-term power supply contracts for data centers continuing to be secured. The energy sector as a whole has been repriced from a traditional defensive position to a core beneficiary of AI infrastructure.

Three: Storage — the most overlooked but longest-lasting benefit

The core logic behind favorable storage is simple: AI is not a one-time call—it is fundamentally a system that continuously ingests, accumulates, and invokes data.

Training requires reading large amounts of data, checkpoints must be saved during training, inference needs to load models and use caching, and RAG and Agent continuously access knowledge bases, logs, and memory.

This means that what AI brings is not just "more data," but:

· More frequent data read and write operations

· Call more real-time

· More complex to manage

· Greater pressure on migration and caching

Looking further down, the more expensive the GPU, the less it can afford to sit idle, so the industry will increasingly focus on how to deliver data faster and more reliably to compute endpoints.

In other words, as AI continues to develop, storage is no longer just a "warehouse for data," but rather the foundational infrastructure that ensures the entire AI system operates continuously.

Memory chip manufacturers: SK Hynix (000660.KS), Samsung Electronics (005930.KS), Micron Technology (MU)

NAND/SSD/HDD manufacturers: SanDisk (SNDK), Seagate (STX), Western Digital (WDC)

Domestic storage design: GigaDevice, Puyan Semiconductor, Dongxin Semiconductor, Beijing Junzheng, Memblaze, and storage module manufacturers Demingli, Shannon Core, and Jiangbo Long, among others.

Since 2026, the storage sector has been one of the strongest segments in the AI industry chain.

In the U.S. stock market, driven by investments in AI infrastructure and demand for high-capacity storage, Seagate, SanDisk, and Western Digital have all risen significantly this year. According to Reuters at the end of April, Seagate and Western Digital have more than doubled, while SanDisk has risen approximately 350% this year.

The original manufacturers of memory chips have also strengthened, with Micron rising significantly this year, while SK Hynix continues to benefit from HBM shortages and intense demand from major companies for production capacity, reporting a 198% year-over-year increase in first-quarter revenue and a 406% increase in operating profit, further enhancing its profitability.

Final note: First come the chips, then the electricity, and finally storage.

The first wave of AI's realization is chips; the second wave's bottleneck is energy; the third wave of long-term beneficiaries is storage.

Logical reasoning doesn't equate to a comfortable entry point. Structural opportunities exist, but don't blindly chase highs.

What truly matters is not the hype itself, but which layer of the supply chain you're on.

Original link

Click to learn about the open positions at BlockBeats

Welcome to the official BlockBeats community:

Telegram subscription group: https://t.me/theblockbeats

Telegram group: https://t.me/BlockBeats_App

Twitter official account: https://twitter.com/BlockBeatsAsia